Er. Cooper et al., CYTOMEGALOVIRUS AS A COFACTOR IN TRANSMISSION AND PROGRESSION OF PERINATAL HIV-INFECTION, Pediatric AIDS and HIV infection, 3(6), 1992, pp. 302-307
We retrospectively studied a cohort of children with HIV infection to
identify those children with cytomegalovirus (CMV) infection. CMV infe
ction was defined as a positive culture for CMV from urine, throat, or
biopsy specimen, or a positive serologic assay for IgG antibody in ch
ildren greater than 6 months of age. Twenty-five (64%) of 39 HIV-infec
ted children were infected with CMV; 22 (88%) were positive on first c
ulture or serologic assay; 15 (60%) had CMV infection documented prior
to their first birthdays. We compared the cumulative prevalence of CM
V in HIV-infected children (P1 and P2), seroreverters (P3), and in age
-matched HIV-negative controls. Prevalence of CMV infection in HIV-inf
ected children was not significantly different from that in HIV-negati
ve children (25/39 vs. 18/49, p = .07). Seroreverters were less freque
ntly infected with CMV than HIV-infected children (2/26 vs. 25/39, p <
.001). HIV-infected children who were also infected with CMV were mor
e likely to have severe encephalopathy, elevated hepatic enzymes assoc
iated with zidovudine (AZT) administration, and suffer death. CMV dise
ase was identified in 5 HIV-infected children, all with acquisition of
CMV early in life. We conclude that CMV infection occurs in the major
ity of HIV-infected children in our population and adversely affects s
urvival. Most CMV-infected children were infected at initial evaluatio
n, suggesting acquisition early in life.